[{"data":1,"prerenderedAt":123},["ShallowReactive",2],{"navigation":3,"\u002Fblog\u002Frag-vs-ai-wiki-evolution-of-context":25,"\u002Fblog\u002Frag-vs-ai-wiki-evolution-of-context-surround":118},[4],{"title":5,"path":6,"stem":7,"children":8,"page":24},"Blog","\u002Fblog","blog",[9,13,17,21],{"title":10,"path":11,"stem":12},"[object Object]","\u002Fblog\u002Fbanksy-what-not-to-do-with-ai","blog\u002Fbanksy-what-not-to-do-with-ai",{"title":14,"path":15,"stem":16},"Making Sense During the AI Revolution","\u002Fblog\u002Fmaking-sense-during-ai-revolution","blog\u002Fmaking-sense-during-ai-revolution",{"title":18,"path":19,"stem":20},"RAG vs AI Wiki","\u002Fblog\u002Frag-vs-ai-wiki-evolution-of-context","blog\u002Frag-vs-ai-wiki-evolution-of-context",{"title":10,"path":22,"stem":23},"\u002Fblog\u002Fvector-storage-in-rag-choosing-right-backend","blog\u002Fvector-storage-in-rag-choosing-right-backend",false,{"id":26,"title":18,"author":27,"body":28,"date":27,"description":10,"extension":104,"image":27,"meta":105,"minRead":27,"navigation":106,"path":19,"seo":107,"stem":20,"__hash__":117},"blog\u002Fblog\u002Frag-vs-ai-wiki-evolution-of-context.md",null,{"type":29,"value":30,"toc":96},"minimark",[31,35,38,43,56,60,71,75,78,89,93],[32,33,34],"p",{},"A traditional wiki stores information. A RAG system retrieves context dynamically.",[32,36,37],{},"Both solve knowledge access, but with different trade-offs.",[39,40,42],"h2",{"id":41},"ai-wiki-strengths","AI Wiki strengths",[44,45,46,50,53],"ul",{},[47,48,49],"li",{},"Stable editorial workflows",[47,51,52],{},"Clear ownership and approval models",[47,54,55],{},"Strong discoverability for humans",[39,57,59],{"id":58},"rag-strengths","RAG strengths",[44,61,62,65,68],{},[47,63,64],{},"Context assembled at query time",[47,66,67],{},"Better performance on long-tail questions",[47,69,70],{},"Easier integration with operational data sources",[39,72,74],{"id":73},"where-teams-get-stuck","Where teams get stuck",[32,76,77],{},"Many teams try to replace the wiki entirely with RAG. In practice, hybrid models work better:",[44,79,80,83,86],{},[47,81,82],{},"Keep canonical policy\u002Fprocess docs in a wiki.",[47,84,85],{},"Use RAG for synthesis and workflow-level answers.",[47,87,88],{},"Add source citation and freshness indicators.",[39,90,92],{"id":91},"key-design-principle","Key design principle",[32,94,95],{},"Treat context as a product. The ingestion, retrieval quality, and governance model matter as much as model choice.",{"title":97,"searchDepth":98,"depth":98,"links":99},"",2,[100,101,102,103],{"id":41,"depth":98,"text":42},{"id":58,"depth":98,"text":59},{"id":73,"depth":98,"text":74},{"id":91,"depth":98,"text":92},"md",{},true,{"title":18,"description":108},{"The evolution of context management":109,"date":110,"image":111,"minRead":112,"author":113},"when retrieval pipelines beat static knowledge pages, and when they do not.","2026-05-26","\u002Fimages\u002Faws_cloud_infrastructure.jpeg",7,{"name":114,"avatar":115},"Andrés Renaud",{"src":116,"alt":114},"\u002Favatars\u002Fplaceholder.svg","ObvPGPRRF9F7ZuAhjklOtI1mRs1sCue7IIptdfKcqaA",[119,121],{"title":14,"path":15,"stem":16,"description":120,"children":-1},"From vibe coding to harness engineering, a practical perspective on building reliable systems while AI tooling evolves fast.",{"title":10,"path":22,"stem":23,"description":122,"children":-1},"Selecting vector storage is not just a database decision. It is an operational decision.",1780335304825]